Hi,
We are using Asset Share Commons 2.2.2 version and extracting facet count to display alongside filters(Ex: Digital(100), Stationary(150)) on the page. We have added custom predicates and made changes by overriding PagePredicateImpl(com.adobe.aem.commons.assetshare.components.predicates.impl) and QueryBuilderResultsImpl(com.adobe.aem.commons.assetshare.search.results.impl.results) to extract the count. Since asset volume is huge(>50k), we are unable to get the facet count and get an error on the page which says Service Outage: Looks like we are having some issues with our service. We are working hard to bring it online again with the entire page empty.
We have lucene indexes defined under(/oak:index/damAssetLucene-8-custom-10/indexRules/dam:Asset/properties). Also tried the same using query debugger tool ,it shows Running search but with no results . We are able to extract the facet count and the debugger tool also displays results if the dam asset count is close to 7k. If results are more than that it is going for index traversal.
I understand that Facet extraction in QueryBuilder and Oak is slow on large result sets. Are there any alternative ways of doing this without impacting the performance?
Views
Replies
Total Likes
Can you not offload search operation to an external serach engine ? Do you also see the same issue on author instance in search admin filter as well if you use same facet or other predicates as below?
Offloading search is not ideal, we would like to use the built-in capabilities of AEM's search. In our case, we created Predicategroup which would search all
DAM assets tagged with a specific tag and used Bucket API to get the count. In author instance, i will only be able to select tags.
Hi Divya,
I would appreciate if you can let me know if you manage to solve this problem. I have the same issue with almost same amount of assets.
Regards,
Andrijana
Views
Replies
Total Likes
Hi Andrijana,
I just got to know that facet counts perform much better if facetStrategy=oak is used in the query. You can give it a try.
Thanks & Regards,
Divya T
Views
Replies
Total Likes